It’s usually called Trellising in most software, but it also has some other names. Therefore, you may be more familiar with it if I would call it small multiples, lattice, grid or panel. But it is usually the same technique but applied in different ways.
The reason why you would do trellising, or in this case small multiples, is according to Edward Tufte in his book Envisioning Information:
At the heart of quantitative reasoning is a single question: Compared to what? Small multiple designs, multivariate and data bountiful, answer directly by visually enforcing comparisons of changes, of the differences among objects, of the scope of alternatives. For a wide range of problems in data presentation, small multiples are the best design solution.
Or to summarize this is a useful technique for being able to do comparison across charts, most often by using the same scale and axes. But it can also be a useful technique for saving space in a layout of multiple charts by making sure that you align the scale and axes and therefore avoid the need to repeat this information.
The best way to make comparison is to make sure you are comparing apples to apples and not to pears. Within visualization this usually means that you either normalize your data or that you make sure that your scales are standardized. Below here you can see how trellising and standardized scales makes it easier to compare across visualizations.
In this comparison I’m going to look at sales of different product categories across four quarters. As you can see without the standardized scales, the main comparison you can do is to see which quarter is best per product category. If you would take the time to look at each scale, you would be able to figure out which product category is bestselling. But that takes a lot of effort from our users.
But as soon as you standardize the scales you can also see
which product category is best, as well as which quarter within each category,
and across categories. The drawback is that for categories with low sales it's
harder to see which quarter is best.
Space and symmetry
As you can see from both these examples, there is a lot of information that is being repeated and takes up valuable space that we could instead use to give more room to the actual data. So, let’s see what it looks like if we remove some of the repetitions. There are many options on which ones to remove, but in this case, I will just make sure that each row and each column doesn’t have repeating axes.
You could prune it even more, but removing too much information might make it hard to interpret the visualization.
Another use case for trellising is to save space within your dashboards by being able to arrange the different visualizations together that you trellis across. This can be done in some different ways.
Using a column layout, you can help the user arrange the visualizations in a way where you read them from left to right and can see how a value for example change over time.
Another option is to trellis by rows, and thus creating a layout that is very similar to a grouped bar chart.
Lastly, you can also save space by just having one
visualization visible at a time and then have the user step through the trellis
one visualization at a time. This is often referred to as a panel chart.
Different chart examples
What we’ve looked at so far are examples using a bar chart. But trellising can also be done with other chart types, even those that do not have an axis to standardize.
Hopefully this blog post gave you some inspiration on how to
do easy comparison with multiple visualizations, but also how you can better
arrange and work with multiple visualizations within your dashboards to save